PSYC 204 Study Guide - Midterm Guide: Repeated Measures Design, John Tukey, Studentized Range
30 Apr 2014
School
Department
Course
Professor
Document Summary
2 (tests independence of nominal variables: h0: two (nominally scaled) variables are statistically independent (no association), assumptions: A sufficiently large sample size is required (n 20) Average cell frequency should be 5: expected frequency: (row total) * (column total)/ (sample size, df = (r-1)(c-1) Ordinal: kruskall wallis h (rank test, h0: k independent samples come from the same population, df = k-1 when nj 3 & k 3. Uses 2 distribution: h (k-1) = _________, p, if significant, do a pairwise kruskall wallis h test if p<0. 5/k, you get significance, median test. Interval or ratio: one-way anova (compares means of populations, h0 : 1 = 2 = k, assumptions: Independence of observations: f(k-1, n-k)=________ where f = vb/vw, when h0 is rejected, conduct post-hoc tests (scheffe and tukey) to know which pair significantly different: pairwise. Scheffe"s test (for groups of different sizes) most. Conservative: for critical value, need df(b) = k-1 and df(w)= n-k multiply by k-1.
